Interactive volume visualization of general polyhedral grids | JoVE Visualize

Area of Science:

Computer Graphics
Scientific Visualization
Computational Geometry

Background:

Volumetric data visualization often relies on tetrahedralization, which is computationally expensive for complex polyhedral grids.
Non-convex polyhedra with non-planar faces present significant challenges for existing visualization techniques.
The increasing use of complex polyhedral grids in advanced simulation packages necessitates efficient visualization methods.

Purpose of the Study:

To present a novel framework for visualizing volumetric data on complex polyhedral grids.
To introduce a compact, face-based data structure (TSFSL) for mesh representation.
To develop a GPU-based ray-casting algorithm for direct visualization using the TSFSL structure.

Main Methods:

Developed a compact, face-based data structure called two-sided face sequence lists (TSFSL) to represent polyhedral mesh topology.
Implemented a direct GPU-based ray-casting algorithm utilizing the TSFSL representation for efficient rendering.
Employed a mesh decomposition strategy (bricking) to handle large datasets, enabling independent processing of bricks and optimizing memory usage and performance.

Main Results:

The TSFSL data structure compactly represents mesh topology in a 1D array, facilitating efficient 1D texture accesses for visualization.
The GPU-based ray-casting algorithm achieves direct visualization without prior tetrahedralization.
Mesh bricking demonstrates memory savings and performance improvements for large-scale polyhedral datasets.
Feasibility validated through visualization of complex, real-world application data from commercial simulation packages.

Conclusions:

The proposed framework effectively visualizes volumetric data on complex polyhedral grids, overcoming limitations of traditional methods.
The TSFSL data structure and GPU ray-casting offer an efficient and scalable solution for scientific visualization.
This approach is suitable for handling highly complex meshes generated by state-of-the-art simulation software.